Diminishing Marginal Utility on Knowledge Derived vs. Detrimental Mistake

One of my favorite super bowl commercials this year advertises a website for car reviews. I think Hayek might chuckle and nod at its point. This point is loosely that we benefit from the mistakes of others, or restated "the advance and even the preservation of civilization are dependent upon a maximum of opportunity for accidents to happen."
http://www.youtube.com/watch?v=dN5aBhrmibg

I also enjoyed our author's point that "the less likely the opportunity, the more serious will it be to miss it when it arises.", also from the second chapter. All of this seems well and good, but I am inclined to think we are missing some of the story.

An important assumption is that the mistake-education process yields maximum positive results. I challenge this assertion. Hayek does not appear to consider that a mistake may be more detrimental than the learned knowledge. The first oil spill informs our future choices and reduces the likelihood of a repeat, but ONLY if the lesson is applied. Each repeat of the mistake in new settings partially increases our knowledge, but how much is really learned? Is enough learned from that mistake to counter the detriment of these mistakes? Would society not better benefit from rules to limit our detrimental mistakes (assuming well crafted rules).

Nor is it shown to my satisfaction that mistakes would not continue to be repeated to the net detriment of the society. On some points, sure: I know that only I can prevent forest fires and I should always assume the gun is loaded, etc. But on a great many points I simply don't know the idea is established, or I think the evaluation of my society is wrong, or maybe I am not engaged in the outcome and don't find the risk significant enough to stop my action. Then, if I do some detrimental thing that every idiot and their brother knows I shouldn't do, have I truly added to the pool of human knowledge? Have I added significantly enough to counter the losses I have caused?

The inevitable question is how another system would better address my concerns. To that question I have no answer. I prefer the average of one thousand dice rolls to the result of a single role; I prefer the aggregate of society to the whims of a select group. By this I concur with the general aim of Hayek and his Austrian peers. I suppose my difficulty is with the absolute nature of his claims. In our discussion last semester we debated extensively on the merits and flaws of a praxeological approach to economic theory. I still maintain that there are no absolutes, and that we cannot speak in absolutes unless we generalize and obfuscate. While it is helpful to deal in generalities, it is incomplete by definition.